Geomatics glossary

The geomatics glossary was produced on the occasion of the June 2010 symposium and was principally written by Paul Rouet and validated by Hélène Noizet.

Preliminary remarks:

  • Censive (French term): territory ruled by a lord;
  • Features: in a geographic information system a feature comprises information about an entity with a geographic location. A feature is thus defined thanks to a geometry (point, line, polygon, multipoint, etc.) and one or several attributes.

Conceptual Data Model (CDM)

abstract representation of a real-world view that we want to study using an information system. The CDM organises, according to established standards, descriptive properties of objects of this real world into homogeneous object classes (for example censives) each containing object attributes (for example the lord's name). A CDM also clarifies the relationships between these classes. Each feature is individualised by a "key" formed from the values of one or more attributes. A CDM outlines the integrity rules that the attribute values of each feature and each relationship between features must follow. The CDM is an essential component of the design phase of an information system and particularly of a GIS.

Data administration

role of the data administrator who manages data documentation (metadata), sorts and controls data and finally manages their dissemination and particularly manages access rights to the information system: right to consult, modify, update, create or destroy feature classes, right to modify integrity rules, etc. The data administrator also acts as a liaison between designers and producers of the information system and its users. In general we entrust to one or more data administrators the ongoing task of imagining, designing and implementing or any other operation (dissemination mode, interfaces, data structure, etc.) that facilitates and optimises the use of the information system within the limits of the resources at their disposal.

Database Management System (DBMS)

a DBMS is software (for example: Access, Oracle, 4D, PostgreSQL, SQL server, etc.) that physically organises data storage of an information system and optimises and secures access to that data. This software includes a set of operations (create, read, update, destroy) usually structured in a "language" allowing for data handling through queries. Most GIS are coupled with a DBMS.

Digital Terrain Model (DTM)/Digital Elevation Model (DEM)

a DTM consists of a regularly spaced grid that provides terrain elevation for each grid node at ground level. A DEM is similarly formed but gives the altitude of each node at elevation level (building roofs, treetops, road structures, etc.). This relief representation mode replaces the classical representation based on contours. There exists methods for changing from one to the other, but nevertheless with a loss of information during transfer.

Geocoding

action of positioning data with a common analog location into a georeferenced space, for example, a postal address or a regular order along a route. The principle is to interpret the analog location to transform it into geographic coordinates or into a projection system using  spatial reference data that has already been georeferenced (layer of postal addresses, road segments of a route, etc.).

Geographic coordinates

position on the Earth’s surface expressed by a latitude (angle formed by the intersection between the equatorial plane and the line passing through this point and the centre of the Earth) and longitude (angle formed between the meridian plane passing through this point and the plane of the Greenwich meridian). Geographic coordinates are expressed in angular values (degree or grade) which differ from the coordinates in a projection system that are expressed in unit of length.

Geographic Information System (GIS)

whether operating on an individual workstation or in an institutional setting, a GIS is a set of computer resources and data designed for the management of objects with a spatial dimension (features). GIS software differs from drawing software (CAD or Computer Aided Design) due to the fact that features are defined by their coordinates on the surface of the Earth and not by their coordinates on a sheet of paper. Furthermore, with GIS software each feature has attribute properties to describe it and to allow it to connect to other features. A drawing software only works using representation; with GIS software objects are an information medium that can be queried, analysed and represented at will.

Georeferencing

action of locating vector as well as raster type objects based on reference geometry using coordinates in a projection system. Georeferencing a raster can lead to distortions of the original image. Georeferencing of vector objects is either direct or indirect: direct when the object has its own coordinates, indirect when these coordinates are retrieved from those of a reference object to which it is associated. For example, a religious building can be located thanks to the coordinates of a plot of land on which it was identified.

Ground Control Point (GCP)

a GCP is an unambiguously identifiable point for both a raster plan/map to georeference and for an already georeferenced GIS layer.

Isotopy, isocliny, isoaxiality

transmission modalities in the time and space of a lineament due to the presence of a director element or a morphogen (isocliny and isoaxiality are new terms introduced in the archaeo-geography field). Isotopy is the transmission of orientation without change of location. Isocliny is the transmission of orientation but with a displacement in space. Isoaxiality is the transmission of information in the extension of the morphogen.

Layer

features are organised into layers of information that are generally representative of an object class. For example: censives, religious buildings, the road network in 1300 and the Vasserot urban blocks are all different layers. A layer contains one or more objects of the same class and of the same geometric type (point, line or polygon). In a geographic information system the objects of a layer (objects are by definition features) are made up of not only their geometry but also their descriptive attributes. For example, in the censives layer, the lord's name is an attribute of each censive. In the religious buildings layer, their ecclesiastical function is also an attribute.

Linear transformation (affine)

mathematical formula which transforms respectively X and Y coordinates of any point of a raster plan into coordinates in a projection system by using a linear or first degree equation.

Map

thematic representation of a space at a point in time within an explicit scale ratio. A map is itself the result of the processing of one or more layers. For example, a map of 18th century censives may include not only the censives boundaries but also the hydrography, road network, outlines of urban blocks or landmark buildings. All of this information is stored in different layers.

Morphogen

in the context of morphological analysis of planimetric documents, a morphogen is a structuring or perennial element of a landscape which may cause the orientation of new shapes in the more or less long term after its establishment.

Pixel

elementary unit of a raster image. Each pixel has a value that corresponds to the content of the image on each point. This value is expressed in bits (basic unit of information in computing) that can be stored on a digital medium. These bits can also be converted back to an image. The 1-bit encoding allows representation of a simple contrast appearing in black and white (code 0 or 1). The 8-bit, 16-bit or 24-bit encoding allows to widen the color space.

Projection system

the shape of the Earth (geoid) is projected onto a mathematical figure called an ellipsoid as closely as possible to the geoid. A projection system consists of converting geographic coordinates on the ellipsoid (latitude-longitude) into coordinates on a planar surface. In its transformation from a spherical space to a plane reality is inevitably distorted. Therefore, several projection systems exist in order to meet different needs, for example, whether we want to represent the whole of the Earth or a small part of terrestrial space or whether we wish to preserve angles or distances. In the ALPAGE project a Lambert conformal conical projection (LCC) known as Lambert-1 is used for Paris.

Raster

image technically made up of a matrix of pixels ordered in rows and columns. The resolution of the raster image is expressed in number of pixels per linear centimetre or per inch (dpi or "dots per inch"). For example, a digital photograph is typically a raster image. The scan of an ancient map also forms a raster image. A map in raster format may be positioned in a projection system. As a raster is composed of pixels which have their own dimensions a change in scaling (zoom) will induce a change in the resolution of its display.

Reference geometry

the optimal way to overlay in detail old plans/maps both together or with current plans/maps is to georeference them using the same reference geometry. The most consistent and accurate possible reference geometry is the one adopted. With its own history and drawbacks this reference geometry will be similar to the concept of "virtual terrain" without pretending to act as a substitute for the actual terrain.

The ALPAGE project's adopted reference geometry is the digital plot plan managed by the APUR (2006 version) which is based mainly on DGI's (Direction Générale des Impôts – the equivalent of the Tax Office) former digital cadastre maintained by the APUR. The method used in the context of this project is to take advantage of the presence of the same features on successive plans allowing for the use of corresponding Ground Control Points (GCPs) either directly (for example, the same building is shown the Vasserot plan and on the geometric reference plan), or indirectly, (for example, when a Vasserot plan no longer has any object in common with the reference plan). If an indirect approach is needed a search for an intermediate plan would be made (for example, a plan of expropriation under the Second Empire) which would then allow for the definition of GCPs both in the reference and the Vasserot plans.

Resolution

number of pixels of a raster image per unit of length. Another way to express the resolution, often used to characterise scanned plans/maps or orthophoto-plans regardless of their scale, is to indicate the ground pixel size. Note: the human eye perceives on paper a resolution of up to around 300 dots per inch (dpi) or around 12 points per millimetre. A computer screen offers a maximum resolution of around 75 to 100 dots per inch.

Root Mean Square (RMS)

this indicator is used in GIS, for example, when a scanned plan or map (raster image) is georeferenced comparatively to one or more already georeferenced GIS data layers. This operation consists of identifying simultaneously on the raster image to georeference and on the already georeferenced GIS layer some specific points (landmarks, building corners, road intersections) called Ground Control Points (GCPs). Mathematically, a minimum of two GCPs are sufficient to georeferenceraster image. However, the deformations in this raster image due to the stretching or contraction it initially contains or simply the difficulty in positioning with certainty each GCP requires, as much as possible, a large number of well spread out points to be selected on the image (3 to 8 or 10 points, beyond this the gain in precision are usually minuscule). Therefore, an algorithm is used that will minimise the distance between the position of each GCP on the original raster image in coordinates and its position on the already georeferenced GIS layer. This distance indicator is the square root of the sum of each square distance or the RMS.

Scanning

method of electronic scanning made up of a light source and a sensor. The light emission is reflected by the document and the reflectance is recorded by the sensor in a raster format called, in this case, a scan.

Spatial Adjustment

process consisting of georeferencing vector data originating from the vectorisation of an already georeferenced raster. The parameters used for this georeferencing process are identical to those used for the georeferencing of the raster. This process, which has been automated in the ALPAGE project, allows an exact overlapping of the georeferenced raster and the newly georeferenced vector data derived from this raster.

Spatial Analysis

operation characteristic of GIS software that determines the spatial relationships between features belonging to one or more layers: inclusion, intersection, exclusion, buffer distance, juxtaposition, etc. We distinguish the analysis of spatial relationships between polygons, between lines, between points and between points and polygons, between lines and polygons, etc.

Topology

property of vector objects. There is a distinction between network topology and surface topology. Network topology for a set of lines consists of creating a planar or non-planar graph made up of edges and nodes. Surface topology is associated with the properties of juxtaposition, superposition, inclusion and exclusion of surface objects in relation to each other. In order to adhere to topological rules, controls of each layer of information or even controls between layers must be made depending on the integrity rules for each of these layers.

Vector

in the domain of GIS a "vector layer" is a mathematical representation of features situated in space with each feature represented as coordinates in a geographical georeferencing system. In a GIS, features can be represented depending on the particular case either by points, lines or polygons. Here we are concerned with conventions: for example, a religious building can only be represented by a point, visible as a symbol to indicate in a simple way its approximate location in space, but if one knows its exact location on the Earth, we can also draw it as a polygon. A vector allows scaling without change of resolution because every vector object is reduced mathematically to sets of points without area. However, this can also be misleading as a vector layer is limited by its own topographical precision. Therefore, resolution should not be confused with precision.

Vectorisation

process of allowing for the creation of vector objects from a raster image. We distinguish manual vectorisation, which requires an operator to interpret the source image themselves, and automatic vectorisation, which is based on the use of shape recognition algorithms which have been previously designed and calibrated according to the source documents and the objectives sought.

Webmapping

putting online of the geographic information system and more widely maps in order to allow for their dissemination throughout the web (internet network). In general, the tools to handle geographic data which are available through the web are a more or less extended subset of the access functions managed by the data administrator. Access rights to functions accessible through the web are also managed by the data administrator, assisted by a webmaster who specifically manages the interfaces and also more generally manages the web services of the website from which the information system is accessed.